{"paper_id":"2509.05911","title":"Deep Learning Option Pricing with Market Implied Volatility Surfaces","authors":["Lijie Ding","Egang Lu","Kin Cheung"],"paper_date":"2025-09-07","paper_type":"empirical","paper_url":"https://arxiv.org/pdf/2509.05911v1","page_url":"https://thequant.space/flowcharts/deep-learning-option-pricing-with-market-implied-volatility/","scores":{"empirical_rigor":8,"math_complexity":7,"quadrant":"Holy Grail"},"proposes":{"status":"available","statement":"The paper proposes a deep learning framework that uses a variational autoencoder to compress market-implied volatility surfaces and a multilayer perceptron to price American puts and arithmetic Asian options in a single forward pass.","quote":"We present a deep learning framework for pricing options based on market-implied volatility surfaces.","summary":"We present a deep learning framework for pricing options based on market-implied volatility surfaces. Using end-of-day S&P 500 index options quotes from 2018-2023, we construct arbitrage-free volatility surfaces and generate training data for American puts and arithmetic Asian options using…","strategy":null},"assumptions":{"status":"available","items":[{"assumption":"The volatility surfaces constructed from end‑of‑day quotes are arbitrage‑free.","type":"market","stated":"explicit","quote":"we construct arbitrage-free volatility surfaces"},{"assumption":"Ground‑truth option values are generated by QuantLib, assuming its numerical methods are accurate.","type":"model","stated":"explicit","quote":"ground truth valuations are obtained from Quantlib"},{"assumption":"Training and testing data are independent because no volatility surfaces from the test set are used in training.","type":"data","stated":"explicit","quote":"The training and testing sets are independent, i.e., the neural network was not trained with any volatility surfaces from the set of testing pricing data."},{"assumption":"Option pricing inputs are limited to end‑of‑day quotes (no intraday dynamics).","type":"market","stated":"implied","quote":"Using end-of-day S&P 500 index options quotes from 2018-2023"}]},"datasets":{"status":"available","items":[{"name":"S&P 500 index options quotes","provider":null,"kind":"options","access":"public","period":"2018-2023","universe":"SPX","quote":"Using end-of-day S&P 500 index options quotes from 2018-2023, we construct arbitrage-free volatility surfaces"},{"name":"Generated American put and arithmetic Asian option valuations","provider":"QuantLib","kind":"synthetic","access":"public","period":null,"universe":null,"quote":"generate training data for American puts and arithmetic Asian options using QuantLib"}],"known":[],"sample":{"start":null,"end":null,"market":null,"asset_class":null,"frequency":null},"data_resolution":"daily"},"transaction_costs":{"status":"not_applicable","modeled":"n/a","headline_net_of_costs":null,"detail":null,"quote":null,"source":"front_matter","rule":"R3","candidates":{"front_matter":"n/a","evidence_card":"n/a","evidence_card_quoted":false},"meaning":"the paper reports no trading or execution result"},"results":{"status":"available","evaluation_type":"pricing-calibration","findings":[{"finding":"The trained pricer achieves high accuracy across American and Asian options, with errors mainly near long maturities and at‑the‑money strikes.","kind":"predictive-accuracy","quote":"The trained pricer achieves high accuracy across American and Asian options, with prediction errors concentrated primarily near long maturities and at-the-money strikes"},{"finding":"Neural‑network predicted prices align well with ground‑truth QuantLib prices, distributing nicely around the diagonal of perfect matches.","kind":"performance","quote":"the neural network predicted prices align well with the ground truth prices, with all of the predicted prices distributing nicely around the diagonal line of perfect matches"},{"finding":"The VAE can accurately reconstruct full volatility surfaces from a 10‑dimensional latent space.","kind":"theoretical","quote":"After we have confirmed that the VAE is able to successfully and accurately reconstruct full volatility surfaces from the latent space"}],"reported_performance":null,"out_of_sample":{"method":"holdout","headline_out_of_sample":null}},"independent_evidence":{"status":"none_found","supports":[],"contradicts":[],"qualifies":[],"compared_with":20,"citations":{"cited_by_in_corpus":0,"by_independent_authors":0,"how_cited":{"used_as_baseline":0,"uses_method_or_data":0,"extends":0,"discusses_findings":0,"mentions":0,"unclassified":0},"independent_papers":[],"scope":"Citations between papers in this corpus only; citations from outside it are not counted."}},"implementation":{"status":"available","public":true,"code_url":"https://github.com/ljding94/VAE_pricing","note":null},"replication":{"status":"none_found","card":null,"note":"No replication of this paper has been run by thequant.space yet."},"extraction":{"text_source":"fulltext","model":"openai/gpt-oss-120b","version":1,"extracted_at":"2026-11-14T17:27:33Z"},"provenance":{"claims":"Everything here is what the authors state in the paper. Nothing was re-run or reproduced by the platform.","quotes":"Assumptions, datasets, cost detail and findings were extracted by a language model. Every quote is verbatim paper text. Where an item has `support`, \"full\" means every figure in it appears in the quoted sentences and \"partial\" means a figure is stated elsewhere in the paper; items with a figure found nowhere in the paper were discarded.","datasets_known":"datasets.known lists well-known vendors and public datasets by a stable id, found by a dictionary over the full text where the sentence says the data is used; it is independent of the model and covers only names in the dictionary.","flags":"transaction_costs.modeled, results.evaluation_type and results.out_of_sample.method are platform classifications (structured-decision models); see https://thequant.space/score-guide/#penalty-flags","comparisons":"supports / contradicts / qualifies are a language model's reading of two papers' quoted findings, restricted to closely related papers that share no author. They are pointers to read, not verified replications, and coverage is partial.","implementation":"code_url is a repository the paper itself presents as its own (rule-scored, ambiguous cases judged by a model), checked for liveness when it was recorded."},"docs":"https://thequant.space/agents/"}